Hi All,
I am planning to design a small module that would make use of object oriented concepts of Perl.
Since its my first attempt in the field of Oops, I would request Perl experts to review the code design as and when I would share the new piece of development for this module.
Basic idea of the module: In many projects, qa needs to validate database tables by executing many sql queries (test cases). The category of these sql queries could be:
1. Count sql queries where in count returned by each sql query should be exactly equal to the known count
2. Count sql queries where in count returned by each sql query should be more than known count.
3. Count sql queries where in count returned by each sql query should be less than known count.
4. Minus sql queries (sql1 minus sql2) where in expected output is 'no rows returned' meaning data set of the the sql queries is exactly same.
The test cases i.e. sql queries would be mentioned in configuration file (called cfg) - 1 sql query per line.
The configuration file would be given for processing to appropriate method. This method would then execute all the test cases and generate the detailed results in the result directory specified by the user.
I have attached initial raw draft of the module and also a test file to test the module.
PS: Its just the first version and with only 1 method. I plan to implement the remaining 5-6 methods gradually.
The reason I want to get it reviewed is because if it turns out to be success I would love to release it to CPAN
Please feel free to share your valuable suggestions.
Main Module
# Author: Parag Kalra, paragkalra@gmail.com
# Licensed under GNU Public License - http://www.gnu.org/licenses/gpl.
+html
package QA::Automation;
use strict;
use warnings;
=head1 NAME
QA::Automation - This module is used to execute Oracle database test c
+ases (sql queries) and validate the automation suite
=head1 DESCRIPTION
This module provides tools (methods) Automate SQL query execution agai
+nst Oracle database
To start with, user needs to create the object by providing database c
+onnection details, result directory and some other
non-mandatory parameters.
As of now plan is to provide following methods:
1. create_cfg_from_uncommented_sql : This method creates cfg (configur
+ation file) from an uncommented sql query.
This cfg forms part of the automation suite that would be executed by
+appropriate method
2. execute_exact_count_sqls : This Method would execute all the count
+sqls mentioned in the cfg and compare the
returned count against the count mentioned for the respective sql quer
+y.
All the test cases where mismatch would be found in the count, would b
+e marked as FAILED or else PASSED
Once all the test cases (sql queries) are executed, a detailed result
+summary would be created in the result
directory specified by the user.
3. execute_minus_match_sqls : This method would execute all the minus
+sqls mentioned in the cfg and would check
if any data is returned or not. If returned then test cases would be m
+arked as FAILED or else PASSED
The intention is to check that data set of minus separated queries is
+same.
Once all the test cases (sql queries) are executed, a detailed result
+summary would be created in the result
directory specified by the user.
Any many more methods are in pipe line
=head1 SYNOPSIS
use QA::Automation;
my $qa_atm = QA::Automation -> new (
uname => "scott",
passwd => "tiger",
+
sid => "orcl",
result => "home/my
+/dir",
ORA_HOME => 1,
DETAILED_RESULT =>
+ 1,
CREATE_RESULT_DIR
+=> 1,
);
=head1 REQUIRES
DBI
DBD::Oracle
File::Path
=cut
# Required modules
use DBI;
use File::Path;
=head1 CONSTRUCTOR AND STARTUP
=head2 new
=over 4
This method creates the object required for automation of test cas
+e execution.
It also creates the result directory if it does not exists and use
+r wants to create it.
=back
=cut
sub new {
my $class = shift;
my %args = @_;
my $self = \%args;
my $result_dir = $self->{'result'};
# if the directory doesn't exists create then create it
if ((! -d $result_dir) && ($self->{'CREATE_RESULT_DIR'})){
mkpath ($result_dir);
}
bless $self, $class or die "Can't bless $class: $!";
return $self;
}
=head1 METHODS
=head2 create_cfg_from_uncommented_sql
=over 4
This method prepares the cfg from the sql file provided as input b
+y the user.
Pre-requisites for this method:
1. All the sql queries mentioned under the sql file should end wit
+h a semi-colon (;)
2. There should be no comments in the sql file
=back
=cut
sub create_cfg_from_uncommented_sql {
my $self = shift;
my $sql = shift;
# Declaring variables
local $/ = ';'; # Setting the IFS to a semicolon
my $input_file;
my $output_file = "$self->{'result'}/SQL.out";
# Checking the existence of the input file
if ( ! defined $sql){
die "Input SQL file not provided - Usage: qa_atm_obj->create_c
+fg_from_uncommented_sql(SQL_File>\n";
} else {
$input_file = $sql;
}
# Processing input/output files
open my $input_fh, '<', $input_file or die "Could not open the inp
+ut file - $input_file\n";
open my $output_fh, '>', $output_file or die "Could not create the
+ output file - $output_file\n";
# Looping through the file and removing all the white space and ap
+pending a new line at the end
while (<$input_fh>){
$_ =~ s/\s+/ /g;
$_ =~ s/^\s+//;
$_ =~ s/\s$//;
print $output_fh "$_\n";
}
# Closing the files
close $input_fh or print "Could not close the file - $input_file\n
+";
close $output_fh or print "Could not close the file - $output_file
+\n";
}
=head1 AUTHOR
Parag Kalra, paragkalra@gmail.com
=cut
1;
Test Script
use strict;
use warnings;
use QA::Automation;
my $qa_atm = QA::Automation->new (
uname => "scott",
passwd => "tiger",
sid => "orcl",
result => "home/my
+/dir",
ORA_HOME => 1,
DETAILED_RESULT =>
+ 1,
CREATE_RESULT_DIR
+=> 1,
);
my @tmp_arr = %{$qa_atm};
$qa_atm->create_cfg_from_uncommented_sql('id_pwd.sql');
